Age | Commit message (Collapse) | Author | |
---|---|---|---|
2000-09-28 | sync | Aaron Campbell | |
2000-09-28 | Add product ID for new CS4281 audio device found in IBM ThinkPad X20. | Aaron Campbell | |
2000-09-28 | If this is a Davicom DM9102A and we're enabling the homePNA link, force | Aaron Campbell | |
dc_link to 1 and don't activate the tick routine. Without this, dc_start() always thinks the link is down and never transmits in homePNA mode; from FreeBSD. | |||
2000-09-28 | Make mouse support friendly with non-25line mode. Thanks mickey@ for reporting | Aaron Campbell | |
the problems. | |||
2000-09-28 | Make sure a Copybuffer has been allocated successfully before we start doing | Aaron Campbell | |
things with it. | |||
2000-09-28 | When allocating the unallocated file descriptors 0, 1 and 2 for suid execs, | Artur Grabowski | |
don't do it by doing namei on /dev/null. The vnode for the executed file is locked and we had a race where other processes could lock the parent directories up to the root. When the executing process did the lookup on /dev/null it could deadlock on the root vnode while still holding the lock on the executed vnode. Also, it's really bad idea to depend on certain filesystem layout inside the kernel. Now we get the null device vnode by cdevvp(getnulldev(), ... Thanks to Matrin Portmann <map@infinitum.ch> for providing the (large) ktrace that allowed me to track this down. Fixes 1369. | |||
2000-09-27 | replace MALLOC/FREE w/ malloc/free for non-constant-sized memory ↵ | Michael Shalayeff | |
allocations; art@ ok | |||
2000-09-27 | Minimal optimization. | Artur Grabowski | |
2000-09-27 | Fix checking for incoming packets when the remote gateway has been | Angelos D. Keromytis | |
fully specified in the flow. | |||
2000-09-26 | Set the tdbi field in the pkthdr to NULL in MGETHDR -- avoids some | Angelos D. Keromytis | |
panics in Alphas (probably others too). | |||
2000-09-26 | Implement getnulldev that returns the dev_t for "/dev/null". | Artur Grabowski | |
2000-09-26 | Don't use MALLOC/FREE on variable sized allocations. | Artur Grabowski | |
2000-09-26 | Update to previous fix on ICMP messages coming on unnumbered | Angelos D. Keromytis | |
interfaces: rather than picking *some* non-loopback IP address, do a routing lookup and use as source IP address the address of the outgoing interface. A nice side effect of this is that ICMPs generated as a result of packets received over IPsec will, in the common case, end up going back over IPsec (depends on what the SPD looks like of course). Thanks to fcusack@fcusack.com for testing and commenting on this. | |||
2000-09-25 | do not do queue locking in get/put_ccb, since calling blocks | Michael Shalayeff | |
have gotten it already locked appropriately. only try to satiate the command queue if there was a command interrupt, it seems it's quite well-behaving this way. dispose allocated resources if twe_attach() fails. | |||
2000-09-25 | Support the i82562; from FreeBSD. Apparently nothing more needed than to match | Aaron Campbell | |
its PCI device ID. | |||
2000-09-25 | sync | Aaron Campbell | |
2000-09-25 | Add Intel i82562 PCI device ID. This is the Ethernet controller found on the | Aaron Campbell | |
Intel 815E motherboard; from FreeBSD. | |||
2000-09-25 | once again bios does not return a correct segment limits for the | Michael Shalayeff | |
pcibios segments. we work this around as we used to do in apm (; round segment size to the next 64k boundary. | |||
2000-09-25 | on expiry of pmtu route, retry higher mtu. okay angelos@ | Niels Provos | |
2000-09-24 | For the Toshiba TOPIC95B CardBus controller, we must clear the socket and slot | Aaron Campbell | |
control registers at shutdown else we hang. Fixes it on my laptop, should help out espie@ on his Toshiba system too. | |||
2000-09-24 | prevent memory leak in fdalloc; fix by greg@nest.cx | Niels Provos | |
2000-09-23 | Angelos you forgot this one !! | Chris Cappuccio | |
2000-09-22 | sync | Aaron Campbell | |
2000-09-22 | Add Conexant modem | Aaron Campbell | |
2000-09-22 | sync | Aaron Campbell | |
2000-09-22 | shorten | Aaron Campbell | |
2000-09-22 | Move the PI_MAGIC define outside the INET6 ifdef block (doh!) | Angelos D. Keromytis | |
2000-09-22 | - Introduce character classes into the console mouse functionality. Lifted from | Aaron Campbell | |
xterm sources. Basically, base word boundaries by character classes rather than space characters only. - On any console logout, clear the contents of the Copybuffer. (This will have to be documented with an explanation as to why.) Avoids the ability to paste the copybuffer at the ``login:'' prompt after someone has logged out. - Permit copy selection extension under any circumstance as long as the button is held down. However, double-click (word selection) does not yet extend on word boundaries like xterms do. - Sanity checks in current_uid(). - From Jean-Baptiste.Marchand@epita.fr | |||
2000-09-22 | sync | Aaron Campbell | |
2000-09-22 | Sony Memory Stick Media Slot | Aaron Campbell | |
2000-09-22 | Add prototypes for ipx_addr() and ipx_ntoa(); mickey@ OK'd | Todd C. Miller | |
2000-09-22 | this follows the dox, saying that %ds should be loaded w/ | Michael Shalayeff | |
the proper segment address/selector as returned from bios32_service(). both %cs and %ds have to be the same value (hmm, same descs?), sincw some bioses are pic, and care only about offsets against the segs. huge, enormous, 10X to brad@ for providing an incredibly usefull testbed and invaluable time of his, for resets and incouraging comments (; | |||
2000-09-22 | fix my bug dating back to february the 14th of 1998, | Michael Shalayeff | |
when those wildcard interfaces came up, which were usefull at the times. on the other hand here it is, one cannot bind to the broadcast address, and angelos says ok. | |||
2000-09-21 | Bad mickey doesn't know how to spell ``openings'' | Miod Vallat | |
2000-09-21 | do not limit openings to defined SDOUTSTANDING value of 4, since many ↵ | Michael Shalayeff | |
controllers can do more; presumably all the cotrollers initialize openings to the right value; also replace some spaces w/ tabs | |||
2000-09-21 | initialize sc_link.openings to it's (currently) default value of 4 | Michael Shalayeff | |
2000-09-21 | calculate maxopd at the right place | Niels Provos | |
2000-09-21 | oops forgot the other case in the pmsintr for add_mouse_randomness() call | Michael Shalayeff | |
2000-09-21 | use volatile's on all structures shared between cpu and hifn | Jason Wright | |
kill hifn_command_buf_data_t (unnecessary for a long time now) | |||
2000-09-21 | style | Jason Wright | |
2000-09-21 | Patch from FreeBSD (wpaul): | Jason Wright | |
Fix the tl driver so it doesn't run off the end of the RX descriptor list and panic in out of mbufs condition. | |||
2000-09-21 | Correctly handle srcid/dstid. | Angelos D. Keromytis | |
2000-09-20 | check out the bios device name we are probed w/ to not probe on apm | Michael Shalayeff | |
2000-09-20 | Add IDENTITY payloads to flow establishment (and cleanup accordingly) | Angelos D. Keromytis | |
-- this will address one of itojun's question on how are IDs for IKE to be determined (need to add support for this to ipsecadm). | |||
2000-09-20 | Don't use LOOPBACK addresses when trying to determine source address | Angelos D. Keromytis | |
to use in locally-generated ICMP messages (thanks to fcusack@fcusack.com) | |||
2000-09-20 | support 64bit BARs from NetBSD | Niklas Hallqvist | |
2000-09-20 | Off by one in a validity check | Niklas Hallqvist | |
2000-09-20 | Extend the region in which we allow mapping unmapped PCI devices. | Niklas Hallqvist | |
The former code only allowed one fixed up device mapping. | |||
2000-09-20 | remove unused code | Niels Provos | |
2000-09-20 | correctly calculate mss | Niels Provos | |